Monitoring method and system based on eigen probability density function

Abstract

The invention provides a monitoring method and system based on an intrinsic probability density function. The monitoring method comprises the following steps: acquiring original data; preprocessing the data to obtain a data time sequence to be analyzed; a step of applying the EEMD to the data time sequence and decomposing the data time sequence into E-IMFs; and an iPDF calculation step, wherein a plurality of iPDFs are obtained. At least a portion of the plurality of iPDFs is used as an output of the monitoring method.

Description

Technical Field

[0001] The present application relates to the field of medical parameter monitoring, and particularly to a monitoring method and system based on an intrinsic probability density function. Background Art

[0002] Anesthesia is a unique medical procedure. It has no therapeutic value in itself, but is indispensable in medical interventions, ranging from life-saving surgeries to routine invasive medical examinations. Methods for measuring the depth of anesthesia (DOA) are crucial in surgical operations and play a good auxiliary role in the surgical process.

[0003] Existing methods for detecting the depth of anesthesia include continuous electrocardiogram (ECG) monitoring, continuous pulse oximetry monitoring (SpO2), blood pressure monitoring, anesthetic concentration monitoring, carbon dioxide monitoring, body temperature measurement, electroencephalogram (EEG) monitoring, etc.; among them, most devices are designed to monitor important physiological indicators mainly related to cardiovascular or pulmonary functions, while the electroencephalogram (EEG) is a device specifically used to monitor the depth of anesthesia through consciousness assessment and is the only device designed to monitor the dynamic function of the brain to determine the patient's level of consciousness; it is the basis for many objective measurements of the depth of anesthesia. Currently, methods for electroencephalogram (EEG) monitoring include compressed spectral array (CSA), bispectral index (BIS), methods such as index and entropy analysis; however, these existing methods generally have drawbacks such as complex analysis or lack of intuitiveness. Summary of the Invention

[0004] In order to overcome some problems existing in the prior art, the present application provides a monitoring method and system based on an intrinsic probability density function.

[0005] The first aspect of the present application provides a monitoring method based on an intrinsic probability density function, including the following steps:

[0006] Step of obtaining raw data: Collect data as raw data;

[0007] Step of data preprocessing: Preprocess the obtained raw data to obtain a data time series to be analyzed;

[0008] Step of applying EEMD to the data time series: Use the EEMD analysis method to process the data time series and adaptively decompose it into multiple ensemble intrinsic mode functions (E-IMFs); and

[0009] Step of calculating the intrinsic probability density function (step of calculating iPDF):

[0010] Obtain N partial sums by taking the partial sum of N ensemble intrinsic mode functions using formula (1);

[0011]

[0012] Among them, c j (t) is the ensemble empirical mode function (E-IMF). t represents the number of data, N represents the total number of E-IMFs, which is a positive integer, n represents the nth E-IMF among them, and x n (t) represents the sum from the 1st E-IMF to the nth E-IMF;

[0013] Then, calculate the probability density function of the sum of each partial component to obtain N initial intrinsic probability density functions (iPDFs), which constitute an initial set of intrinsic probability density functions (iPDF set);

[0014] Among them, at least a part of the iPDF set (such as 2, 3, 5, or 8 in 1PDF - 8PDF, etc.) is used as the output of the monitoring method.

[0015] In one embodiment, the monitoring method further includes a calculation step of ensemble intrinsic probability density function (eiPDF calculation step): Set multiple white noise sequences with different amplitude ratios, apply them respectively to the step of applying EEMD to the data time series, and perform the calculation step of iPDF to obtain multiple rounds of iPDF sets; Take the ensemble average value of the iPDF sets under multiple rounds as the result to obtain multiple initial ensemble intrinsic probability density functions (eiPDFs), and multiple eiPDFs constitute an initial set of ensemble intrinsic probability density functions (eiPDF set); At least a part of the eiPDF set (such as 2, 3, 5, or 8 in e1 PDF - e8PDF, etc.) is used as the output of the monitoring method, and can replace the iPDF set as the output parameter.

[0016] In one embodiment, in the calculation step of the iPDF, further, subtract the probability density function of the standard normal distribution from the obtained multiple iPDFs respectively to obtain multiple final intrinsic probability density functions (F-iPDFs), and use at least a part of the multiple F-iPDFs as the output of the monitoring method, which can replace the output of the iPDF.

[0017] In one embodiment, in the calculation step of the eiPDF, further, subtract the probability density function of the standard normal distribution from the obtained multiple eiPDFs respectively to obtain multiple final ensemble intrinsic probability density functions (F-eiPDFs), and use at least a part of the multiple F-eiPDFs as the output of the monitoring method, which can replace the output of the eiPDF.

[0018] In one embodiment, in the step of obtaining the original data, data is collected using a sliding window, and the data of each window is used as one of the original data; using the monitoring method, the outputs of each window can be obtained, such as multiple iPDFs, F-iPDFs, eiPDFs, and / or F-eiPDFs.

[0019] In one embodiment, in the step of obtaining the original data, the sampling frequency of the data is not less than 100 Hz.

[0020] In one embodiment, in the data preprocessing step, the data preprocessing includes filtering, outlier removal, and electrical noise removal.

[0021] In one embodiment, the step of applying EEMD to the data time series specifically includes:

[0022] Adding white noise: adding a white noise sequence to the data time series to be analyzed; the amplitude ratio of the white noise is any value in the range of 0.05 - 0.2;

[0023] Using EMD to decompose IMFs: adaptively decomposing the data time series with the added white noise sequence into multiple initial intrinsic mode functions (IMFs) using EMD;

[0024] Continuously repeating the steps of adding white noise and using EMD to decompose IMFs, but each time using a newly randomly generated white noise sequence with the same amplitude ratio; thereby obtaining multiple batches each with multiple IMFs;

[0025] Taking the ensemble average of the multiple IMFs of multiple batches as the decomposition result to obtain the multiple E-IMFs.

[0026] In one embodiment, in the step of applying EEMD to the data time series, the steps of adding white noise and using EMD to decompose IMFs are repeated 10 - 20 times to obtain 10 - 20 batches.

[0027] In one embodiment, in the step of calculating the iPDF, at most 8 of the multiple E-IMFs are selected for calculating the partial component sum; that is, N is a positive integer less than or equal to 8, such as 8, 7, 6, etc.

[0028] In one embodiment, in the step of calculating the eiPDF, the amplitude ratio of the white noise is set to 0.05 - 0.2, and 6 - 16 ensemble trials are taken therein as the multiple white noise sequences with different amplitude ratios.

[0029] In one embodiment, the monitoring method further includes a data output step, wherein the output of each window (at least a part of multiple iPDFs, multiple F-iPDFs, multiple eiPDFs or multiple F-eiPDFs) is displayed as a result on a monitor and archived as a monitoring guide.

[0030] In one embodiment, in the data output step, at least one of the following methods is selected:

[0031] (1) Output all or a part of the multiple iPDFs of each window to obtain an iPDF monitoring graph of each iPDF on the time axis for monitoring guidance;

[0032] (2) Output all or a part of the multiple F-iPDFs of each window to obtain an F-iPDF monitoring graph of each F-iPDF on the time axis for monitoring guidance;

[0033] (3) Output all or a part of the multiple eiPDFs of each window to obtain an eiPDF monitoring graph of each eiPDF on the time axis for monitoring guidance;

[0034] (4) Output all or a part of the multiple F-eiPDFs of each window to obtain an F-eiPDF monitoring graph of each F-eiPDF on the time axis for monitoring guidance;

[0035] (5) Output all or a part of the multiple iPDFs of each window respectively at the standard deviation distribution d to obtain an iPDF monitoring curve of a single iPDF at a single standard deviation distribution d on the time axis for monitoring guidance;

[0036] (6) Output all or a part of the multiple F-iPDFs of each window respectively at the standard deviation distribution d to obtain an F-iPDF monitoring curve of a single F-iPDF at a single standard deviation distribution d on the time axis for monitoring guidance;

[0037] (7) Output all or a part of the multiple eiPDFs of each window respectively at the standard deviation distribution d to obtain an eiPDF monitoring curve of a single eiPDF at a single standard deviation distribution d on the time axis for monitoring guidance;

[0038] (8) Output all or a part of the multiple F-eiPDFs of each window respectively at the standard deviation distribution d to obtain an F-eiPDF monitoring curve of a single F-eiPDF at a single standard deviation distribution d on the time axis for monitoring guidance.

[0039] In one embodiment, both the monitoring graph and the monitoring curve are used for monitoring guidance.

[0040] In one embodiment, in the step of outputting the data, select the third one in the F-eiPDF when d = 0, that is, the F-e3PDF(d = 0) monitoring curve, and the second one in the F-eiPDF when d = -4, that is, the F-e2PDF(d = -4) monitoring curve, as the main results, and use the other monitoring curves as supporting results.

[0041] The second aspect of the present application provides a monitoring system based on the intrinsic probability density function, which adopts the monitoring method described in any one of the previous embodiments. The monitoring system includes a data acquisition module and a data analysis module, where,

[0042] The data acquisition module is configured to collect data so as to obtain the original data;

[0043] The data analysis module includes a windowing module, a data preprocessing module, an EEMD decomposition module, an iPDF calculation module, and a result output module; where,

[0044] The windowing module is configured to window the collected data to obtain the original data of each window;

[0045] The data preprocessing module is configured to filter, remove outliers, and remove electrical noise from the collected original data to obtain the data time series to be analyzed;

[0046] The EEMD decomposition module is configured to adaptively decompose the data time series to be analyzed in the current window into multiple E-IMFs by using the EEMD analysis method;

[0047] The iPDF calculation module is configured to select the first N E-IMFs from the multiple E-IMFs, and use formula (1) to find the partial component sum to obtain N partial component sums,

[0048]

[0049] where, c j (t) is the E-IMF, t represents the number of data, N represents the total number of E-IMFs, which is a positive integer, n represents the nth E-IMF among them, and x n (t) represents the sum from the 1st E-IMF to the nth E-IMF;

[0050] Then, calculate the probability density function of each partial component sum to obtain the iPDF set of the current window;

[0051] The result output module is configured to sort and output the multiple iPDFs of each window.

[0052] In one embodiment, the monitoring system further includes: an eiPDF calculation module configured to set white noise sequences with multiple different amplitude ratios, apply them to an EEMD decomposition module and an iPDF calculation module respectively, and obtain multiple rounds of iPDF sets; take the set average of the iPDF sets under multiple rounds as the result to obtain multiple eiPDFs for the current window; and a result output module configured to sort and output the multiple eiPDFs of each window.

[0053] In one embodiment, the iPDF calculation module is further configured to subtract the probability density function of the standard normal distribution from the obtained multiple iPDFs respectively to obtain multiple F-iPDFs, which constitute an F-iPDF set.

[0054] The result output module is further configured to sort and output the multiple F-iPDFs of each window.

[0055] In one embodiment, the eiPDF calculation module is further configured to subtract the probability density function of the standard normal distribution from the obtained multiple eiPDFs respectively to obtain multiple F-eiPDFs, which constitute an F-eiPDF set.

[0056] The result output module is further configured to sort and output the multiple F-eiPDFs of each window.

[0057] In one embodiment, the monitoring system further includes:

[0058] a data archiving module configured to store data for future reference;

[0059] a data display module configured to display the output result of the result output module on a monitor as a monitoring guide.

[0060] A third aspect of the present application provides an application of the monitoring method based on the intrinsic probability density function described in any of the foregoing embodiments. The monitoring method can be used for medical monitoring, including but not limited to anesthesia depth monitoring and muscle activity monitoring. The original data respectively correspond to electroencephalogram data and electromyogram data.

[0061] A fourth aspect of the present application provides a computer device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, it implements the monitoring method based on the intrinsic probability density function described in any of the foregoing embodiments.

[0062] A fifth aspect of the present application provides a computer-readable storage medium, on which a computer program is stored. When the program is executed by a processor, it implements the monitoring method based on the intrinsic probability density function described in any of the foregoing embodiments.

[0063] The monitoring method provided by at least one embodiment of the present application is based on the intrinsic probability density function or the ensemble intrinsic probability density function, and has the following advantages: (1) The intrinsic probability density function or the ensemble intrinsic probability density function is easier to calculate, and thus can provide a score closer to real time; (2) The calculation of the intrinsic probability density function or the ensemble intrinsic probability density function depends less on the data length than entropy. Entropy depends on the threshold within the data range, which can make the calculation unstable; (3) The measurement of the intrinsic probability density function or the ensemble intrinsic probability density function provides the modulation or interaction of direct physiological and neural information and can be used as a guide for the depth of clinical anesthesia.

[0064] The monitoring method provided by at least one embodiment of the present application is based on the intrinsic probability density function or the ensemble intrinsic probability density function, and can detail the intrinsic probability density attributes of large-amplitude and small-amplitude components, as well as stationary and non-stationary processes; Since the mean and variance values will be dominated by the changes of the components with the largest energy, the global distribution will be dominated by such components, and the effect of the lower-energy components being obscured will be reduced. The intrinsic probability density will improve these situations and enable us to detail the probability density of all scale components. The probability density of some components and that of individual components may be very different, especially when the process involves multi-scale interactions with a non-linear multiplication process.

[0065] The monitoring method provided by at least one embodiment of the present application can be effectively applied to the medical field and other related fields to contribute to the monitoring of data. Description of the Drawings

[0066] Figure 1 is a flowchart of the monitoring method according to an embodiment of the present application;

[0067] Figure 2 is a schematic diagram of collecting electroencephalogram data;

[0068] Figure 3a is the original data of the current window;

[0069] Figure 3b is the data time series to be analyzed after preprocessing;

[0070] Figure 4 is an example of multiple E-IMFs obtained by EEMD decomposition;

[0071] Figure 5 is the eiPDF curve graph when the patient is in a deep anesthesia state;

[0072] Figure 6 is the two-dimensional F-eiPDF graph when the patient is in a deep anesthesia state;

[0073] Figure 7It is the eiPDF curve graph when the patient is in a conscious state;

[0074] Figure 8 It is the two-dimensional F-eiPDF graph when the patient is in a conscious state;

[0075] Figure 9 It is the two-dimensional F-eiPDF monitoring graph obtained by the monitoring method of this embodiment during the entire anesthesia operation process;

[0076] Figure 10 It is the F-eiPDF(d = 0) monitoring curve obtained by the monitoring method of this embodiment during the entire anesthesia operation process;

[0077] Figure 11 It is the F-eiPDF(d = -4) monitoring curve obtained by the monitoring method of this embodiment during the entire anesthesia operation process;

[0078] Figure 12a It is the BIS monitoring curve and the preprocessed electroencephalogram data;

[0079] Figure 12b It is the F-e3PDF(d = 0) monitoring curve;

[0080] Figure 12c It is the F-e2PDF(d = -4) monitoring curve;

[0081] Figure 13 It is the preprocessed electromyogram data of the tremor muscles monitored in Parkinson's patients;

[0082] Figure 14 It is the F-e3PDF(d = 0) monitoring curve obtained by processing the electromyogram data;

[0083] Figure 15 It is the F-e2PDF(d = -4) monitoring curve obtained by processing the electromyogram data;

[0084] Figure 16 It is the two-dimensional F-eiPDF monitoring graph obtained by processing the electromyogram data;

[0085] Figure 17 It is the F-eiPDF(d = 0) monitoring curve obtained by processing the electromyogram data;

[0086] Figure 18 It is the F-eiPDF(d = -4) monitoring curve obtained by processing the electromyogram data;

[0087] Figure 19 It is the connection schematic diagram of the computer device. Specific implementation manner

[0088] To make the objectives, technical solutions and advantages of this application clearer and more understandable, the following describes and explains this application in combination with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not used to limit this application. Based on the embodiments provided in this application, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of this application.

[0089] Obviously, the accompanying drawings in the following description are only some examples or embodiments of this application. For those of ordinary skill in the art, without creative efforts, this application can also be applied to other similar scenarios based on these drawings. In addition, it can also be understood that although the efforts made in such a development process may be complex and lengthy, for those of ordinary skill in the art related to the content disclosed in this application, some design, manufacturing or production changes based on the technical content disclosed in this application are only conventional technical means and should not be understood as the content disclosed in this application being insufficient.

[0090] When "embodiment" is mentioned in this application, it means that the specific features, structures or characteristics described in combination with the embodiment can be included in at least one embodiment of this application. The appearance of this phrase in various positions in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment mutually exclusive with other embodiments. Those of ordinary skill in the art explicitly and implicitly understand that the embodiments described in this application can be combined with other embodiments without conflict.

[0091] Unless otherwise defined, the technical terms or scientific terms involved in this application should be understood in the ordinary sense by those of ordinary skill in the technical field to which this application belongs. The words such as "a", "one", "kind", "the" and the like involved in this application do not indicate a limitation in quantity and can represent a single or plural number. The terms "include", "comprise", "have" and any variations thereof involved in this application are intended to cover non-exclusive inclusion; for example, a process, method, system, product or device that includes a series of steps or modules (units) is not limited to the listed steps or units, but may also include unlisted steps or units, or may also include other steps or units inherent to these processes, methods, products or devices. The terms "connect", "be connected", "couple" and the like involved in this application are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. The "plurality" involved in this application refers to two or more. "And / or" describes the association relationship of associated objects and indicates that three relationships can exist.

[0092] Traditional probability density function (PDF) globally averages and describes data based on the total number of samples. Since it is usually represented by the mean and variance values, for non-stationary processes, it is impossible to use PDF because the mean and variance are meaningless. Even for stationary processes, the final probability density is controlled by components with higher energy, and the characteristics of components with lower energy will be completely masked; importantly, components with lower energy are likely to be the activities of interest. For example, if one is interested in the turbulent motion in the ocean surface layer, the wave motion has much higher energy than the turbulence, so the PDF result cannot reveal any turbulent characteristics; however, the turbulence should still have an inherent probability density structure, which is an inherent characteristic of the underlying turbulent process buried by the higher-energy waves.

[0093] This application proposes the concepts of a new Intrinsic probability density function (iPDF) and Ensemble intrinsic probability density function (eiPDF), and applies them to fields such as medical monitoring, such as the monitoring of anesthetic depth, muscle activity monitoring, and so on. This embodiment mainly takes the monitoring of anesthetic depth as an example for illustration; however, the analysis and processing methods in other fields or directions are the same or similar. The medical community generally believes that the unconscious state is a state in which various parts of the brain can hardly communicate, connect, and interact with each other. Communication, connection, and interaction mean the modulation of brain waves, and the intrinsic probability density function and ensemble intrinsic probability density function provided in this application can quantify the intensity of this interaction, thereby monitoring the anesthetic depth.

[0094] The first embodiment of this application provides an anesthetic depth monitoring method based on the intrinsic probability density function (hereinafter may be simply referred to as the monitoring method). Figure 1 It is a flowchart of the monitoring method according to this embodiment. It should be noted that the steps shown in the flowchart of the monitoring method or in the flowchart of the drawings can be executed in a computer system such as a set of computer-executable instructions. The monitoring method includes the following steps:

[0095] S100 The step of acquiring raw data: As Figure 2 shown, a single electrode deployed on the patient's forehead and a reference electrode near the ear (mastoid process of the temporal bone) can be used to collect the patient's electroencephalogram data (EEG data, unit: μV), thereby acquiring raw data. The sampling frequency of the electrode is generally not less than 100 Hz, for example, it can be 100 Hz, 256 Hz, 500 Hz, and so on.

[0096] To report results close to real-time, windowing can be implemented for the acquisition of EEG data. In this embodiment, a sliding window is adopted; after setting the window length and the sliding step size, the window sliding stage is entered, thereby obtaining the raw data of each window.

[0097] Next, taking specific data as an example, a more detailed description will be given for better understanding; however, the protection scope of this application is not limited thereto. Set the sampling frequency of the electrode to 100 Hz, the window length to 60 s, and the sliding step size to 1 s; after turning on the system, start collecting data, collect 100 EEG data every 1 second and transmit them to the computer; when the computer has received 60 s of EEG data, the raw data of the current window is obtained, and this raw data contains 6000 (60 * 100) data. During the window sliding stage, multiple raw data can be obtained in sequence. For example, since the window sliding stage is entered, after starting to collect EEG data, first, 6000 data from the 1st second to the 60th second are obtained as the first raw data of the first window; after the window moves 1 s, 6000 data from the 2nd second to the 61st second can be obtained as the second raw data of the second window; after the window moves 1 s again, 6000 data from the 3rd second to the 62nd second can be obtained as the third raw data of the third window, …, and so on. In this way, multiple raw data of each window can be obtained in sequence during the window sliding stage. According to this method, numerous raw data of each window from the start to the end of the operation can be obtained. The first 59 s of the subsequent window are the same as the last 59 s of the previous window, so that results can be reported close to real-time.

[0098] For the acquisition of other monitored data, corresponding acquisition devices and means can be used. For example, for the acquisition of muscle activity data, electromyography devices can be used for acquisition, and the sampling frequency of electromyography data is generally above 1000 Hz.

[0099] S200 Data preprocessing step: Preprocess the raw data of the currently obtained window to obtain the data time series to be analyzed for the current window.

[0100] It should be noted that when analyzing data subsequently, only the data of the current window is taken as an example for analysis; the data of other windows are processed in the same way.

[0101] Since the amplitude of EEG data of a typical adult scalp is only 10 μV to 100 μV, which is relatively small, but the obtained raw data is full of signal interference with relatively large amplitudes, such as interference from sources like blinking, teeth grinding, body movement, touching EEG electrodes and nearby power supply devices, and there is also drift. Therefore, the raw data needs to be preprocessed.

[0102] The data preprocessing includes filtering, outlier removal, and electrical noise removal.

[0103] Taking specific data as an example, the sampling frequency of the original data is 100 Hz, and 6000 pieces of original data are obtained in the current window. The filter is set to high-pass 0.5 Hz and low-pass 48 Hz. Thus, the high-pass filter removes frequencies below 0.5 Hz, and the low-pass filter removes frequencies above 48 Hz, simultaneously achieving the removal of electrical noise (50 Hz power frequency). In addition, to eliminate the interference of other devices or movements in the operating room, all outliers are deleted. Specifically, any value greater than the mean plus 5 times the standard deviation of the absolute value of the filtered data of the original data in the current window is defined as an outlier. Specifically, the mean and standard deviation of the absolute values of the 6000 filtered data in the current window are calculated. When the absolute value of the filtered data is greater than the mean plus 5 times the standard deviation, it is defined as an outlier and removed; when removing the outlier, the corresponding outlier is filled in with the average of multiple surrounding values. Therefore, the overall data length does not decrease and remains 6000, that is, the data time series to be analyzed obtained after processing still contains 6000 data. Figure 3a is an example of the original data of a certain window, Figure 3b is an example of the data to be analyzed after preprocessing of this window; their abscissas are both time (unit: s), and the ordinates are electroencephalogram data (unit: μV). When the data of a certain window is in the processing state, it is the current window mentioned above.

[0104] For the preprocessing of other monitored data, corresponding methods can be adopted. For example, for electromyogram data, the sampling frequency is set to above 1000 Hz, such as 1000 Hz, 2000 Hz, etc. Here it is set to 1000 Hz, the window is set to 10 seconds (that is, the window length is 10 s), and the sliding step is 1 s, so 10000 pieces of original data are obtained. For the preprocessing of 10000 pieces of original data, high-pass filtering of 0.5 Hz and low-pass filtering of 450 Hz can be performed; in addition, to remove the interference of power frequency 50 Hz and its harmonics, 50 Hz notch filtering and 100 Hz notch filtering can also be performed.

[0105] Steps of applying EEMD to the data time series: Use the EEMD analysis method to process the data time series to be analyzed in the current window, and adaptively decompose it into multiple ensemble empirical mode functions E-IMFs (s represents a complex number, multiple).

[0106] In the prior art, Empirical Mode Decomposition (EMD, Huang et al., 1998) is an adaptive method designed to analyze non-linear and non-stationary data. It can extract Intrinsic Mode Functions (IMFs) from any data and has the following good characteristics: they are all zero-mean, symmetric about zero, and binary narrow-band; therefore, the IMFs provide a compact support for the distribution of the data from the trend scale to the whole. To ensure the stability of the EMD method, an ensemble method with different added noises can also be used to assist the decomposition (Wu and Huang, 2009), which is called Ensemble Empirical Mode Decomposition (EEMD). EEMD can be implemented through the following steps: (1) Add a white noise sequence to the target data sequence to be analyzed; (2) Decompose the target data sequence with added white noise into IMFs using EMD; (3) Continuously repeat steps (1) and (2), but use different white noise sequences each time; (4) Take the ensemble average of the IMFs obtained from multiple trials as the final decomposition result to obtain Ensemble Intrinsic Mode Functions E-IMFs.

[0107] In this embodiment, step S300 specifically includes:

[0108] S301 Add white noise: Add a white noise sequence to the data time series to be analyzed in the current window. This white noise sequence is randomly generated and has the same length as the data time series. The amplitude size of the white noise is determined according to the amplitude size of the data to be analyzed, generally any value in 0.05 - 0.2 of the amplitude of the data time series, that is, the amplitude ratio of the white noise is selected as any value in 0.05 - 0.2. If it is too small, the effect is not obvious; if it is too large, it will interfere with the original data.

[0109] S302 Decompose IMFs using EMD: Decompose the data time series with the added white noise sequence into multiple initial Intrinsic Mode Functions IMFs using EMD adaptively to form a set of initial Intrinsic Mode Functions (IMFs set).

[0110] S303 Continuously repeat S301 and S302, but use a newly randomly generated white noise sequence with the same amplitude ratio each time; thus, multiple batches with IMFs sets are obtained respectively. Generally, repeat 10 - 20 times to obtain 10 - 20 batches, and each batch has an IMFs set containing multiple IMFs.

[0111] S304 Take the ensemble average of the IMFs sets of multiple batches as the decomposition result to obtain multiple Ensemble Intrinsic Mode Functions E-IMFs, forming a set of Ensemble Intrinsic Mode Functions E-IMFs (E-IMFs set).

[0112] Among them, in step S303, the data is processed by using a method of multiple cyclic decompositions with different white noises each time. Such a processing method can not only make up for the deficiencies of EMD, but also reduce the influence of white noise on the final data.

[0113] The described set of IMFs contains multiple IMFs, which can be IMF1, IMF2, IMF3, etc. The described set of E-IMFs contains multiple E-IMFs, which can be E-IMF1, E-IMF2, E-IMF3, etc. As is well known to those skilled in the art: when calculating the ensemble average, it is the result obtained by averaging each element in each set separately. For example, in step S304, the ensemble average of multiple batches of sets of IMFs is obtained by averaging IMF1 in each batch, IMF2 in each batch, …, so as to obtain a set of E-IMFs containing these averages.

[0114] Both the set of IMFs and the set of E-IMFs can be used in the subsequent calculation steps of iPDF. Since the set of E-IMFs has undergone multiple iterations relative to the set of IMFs, the set of E-IMFs is more preferred. However, it should be understood that when describing the decomposition result of EEMD, in the absence of the set of E-IMFs, its result can also be considered as the set of IMFs, that is, the set of IMFs is regarded as the described multiple set intrinsic mode functions. At this time, EEMD can also be regarded as EMD.

[0115] Continuing with a specific data example. The data time series to be analyzed has 6000 data. A white noise sequence containing 6000 data is randomly generated, and its amplitude is set to 10% of the amplitude of the data time series (amplitude ratio is 0.1). Then, the EEMD analysis method is used for decomposition, and 9 set intrinsic mode functions E-IMFs can be obtained, namely E-IMF1, E-IMF2, E-IMF3, E-IMF4, E-IMF5, E-IMF6, E-IMF7, E-IMF8, E-IMF9, which constitute a set of E-IMFs. Among them, each set intrinsic mode function E-IMF contains 6000 data. As Figure 4 shows an example of setting to add a white noise sequence of 10% and obtaining 9 E-IMFs by EEMD decomposition. Among them, E-IMF1 to E-IMF9 are arranged in order from high frequency to low frequency, which is the result of the adaptive decomposition of EEMD. Since the EMD and EEMD analysis methods are already very mature methods, no more detailed description will be given here.

[0116] Calculation steps of the S400 intrinsic probability density function (iPDF calculation steps): Select up to 8 out of multiple E-IMFs in the current window for calculating the intrinsic probability density function (iPDF) of the current window. Specifically:

[0117] S401 Select the first N E-IMFs in the current window, where N is a positive integer not greater than 8.

[0118] S402 Use formula (1) to find the partial component sums of the selected E-IMFs, obtaining N partial component sums:

[0119]

[0120] where, c j (t) is the ensemble intrinsic mode function (E-IMF), t represents the number of data, N represents the total number of E-IMFs, n represents the nth E-IMF among them, and x n (t) represents the sum from the 1st E-IMF to the nth E-IMF.

[0121] S403 Calculate the probability density functions of each partial component sum, thereby obtaining N initial intrinsic probability density functions (iPDFs) of the current window, forming the set of initial intrinsic probability density functions (iPDF set) of the current window.

[0122] In this S403 step, since the intrinsic mode function is combined with the probability density function, the resulting result is called the intrinsic probability density function (iPDF). Among them, the calculation method of the probability density function (such as formulas, etc.) is a conventional method in this field, and it can be calculated by referring to existing literature (such as Parzen E., “On Estimation of a Probability DensityFunction and Mode”[J], Annals of Mathematical Statistics, 1962, 33(3):1065 - 1076.).

[0123] Continuing with specific data as an example. In step S300, a white noise sequence with an added amplitude ratio of 10% is set, and 9 E-IMFs obtained by EEMD are as Figure 4 shown. Then select E-IMF1, E-IMF2, E-IMF3, E-IMF4, E-IMF5, E-IMF6, E-IMF7, E-IMF8 among them. It should be understood that in some cases, if only 7 E-IMFs are obtained, then all can be selected, not necessarily 8.

[0124] The partial component sums of these 8 E-IMFs are obtained by using formula (1), and 8 partial component sums are obtained respectively. In formula (1), c j (t) is the ensemble intrinsic mode function (E-IMF), t represents 6000 data, N = 8 represents the total number of E-IMFs, n represents the nth E-IMF, which is 1, 2, …, 8, and x n (t) represents the sum from the 1st E-IMF to the nth E-IMF. Each partial component sum obtained by formula (1) still contains 6000 data.

[0125] Then, the probability density functions of each partial component sum are calculated respectively. The 8 partial component sums result in 8 intrinsic probability density functions (iPDFs), namely 1PDF, 2PDF, 3PDF, 4PDF, 5PDF, 6PDF, 7PDF, and 8PDF; they form the iPDF set of the current window, and each intrinsic probability density function can be represented by a curve distribution diagram.

[0126] In an embodiment, further, each obtained initial intrinsic probability density function is respectively subtracted by the probability density function of the standard normal distribution to obtain the final intrinsic probability density function (F-iPDF), which constitutes the set of the final intrinsic probability density functions (F-iPDF set) of the current window. The probability density function of the standard normal distribution represents a completely random and disordered state; after subtraction, the obtained F-iPDF makes the characterization of the whole process simpler and more direct compared with the iPDF, facilitating observation and having a more intuitive auxiliary effect. Reference can be made to the specific example part of the final ensemble intrinsic probability density function in the following text for understanding.

[0127] S500 Calculation steps of the ensemble intrinsic probability density function (calculation steps of eiPDF): Set white noise sequences with multiple different amplitude ratios, apply them respectively to step S300 (wherein, when step S303 is repeated, the same amplitude ratio is used), and perform step S400 to obtain multiple rounds of iPDF sets. Take the set average value of the iPDF sets under multiple rounds as the result to obtain multiple initial ensemble intrinsic probability density functions (eiPDFs) of the current window, which constitute the set of the initial ensemble intrinsic probability density functions (eiPDF set) of the current window.

[0128] In one embodiment, the amplitude ratio of white noise is set to 0.05 - 0.2, and 6 - 16 set trials are taken from 0.05 - 0.2. Steps S300 and S400 are respectively performed for each set trial, so as to obtain the iPDF sets for each round. Each iPDF set contains multiple iPDFs. For white noise with different amplitudes, there will be some differences in the E-IMFs obtained by EEMD, and the corresponding iPDF sets will also have some differences; the iPDF sets under multiple trials are averaged to obtain an eiPDF set containing multiple eiPDFs. Its advantage lies in that this is an average value of the set, which is the most likely feature in terms of probability and is also a more stable result.

[0129] In one embodiment, further, the obtained respective initial set intrinsic probability density functions are respectively subtracted by the probability density function of the standard normal distribution to obtain the final set intrinsic probability density function (F-eiPDF), thereby constituting the set of the final set intrinsic probability density functions (F-eiPDF set) of the current window. The probability density function of the standard normal distribution represents a completely random and disordered state; after subtraction, the obtained F-eiPDF makes the characterization of the whole process simpler and more direct compared with the eiPDF, which is convenient for observation and has a more intuitive auxiliary effect.

[0130] Continuing with specific data as an example. Seven amplitude ratios of white noise are taken from 0.05 - 0.2, so that seven set trials can be obtained, which can be used for seven rounds of calculations. In the first round, the amplitude of the white noise sequence is set to 0.1, and step S300 is entered. Through steps S301 - S304, the E-IMF set of the first round (including multiple cyclic iterations of 0.1 white noise) is obtained, and then step S400 is entered. Through steps S401 - S403, the iPDF set of the first round is obtained. Suppose it contains 8 iPDFs (i.e., 1PDF, 2PDF, 3PDF, 4PDF, 5PDF, 6PDF, 7PDF, 8PDF)... and so on... The iPDF sets of all 7 rounds can be obtained.

[0131] The 1PDFs in the iPDF sets of the first to seventh rounds are averaged to obtain the first set intrinsic probability density function e1 PDF, the 2PDFs in the iPDF sets of each round are averaged to obtain the second set intrinsic probability density function e2PDF,..., and the 8PDFs in the iPDF sets of each round are averaged to obtain the eighth set intrinsic probability density function e8PDF, thereby obtaining eight initial set intrinsic probability density functions (eiPDFs), which constitute the eiPDF set. On this basis, the above eight eiPDFs are respectively subtracted by the probability density function of the standard normal distribution, thereby obtaining eight F-eiPDFs, which constitute the F-eiPDF set of the current window.

[0132] Figure 5 andFigure 7 Examples of the eiPDF sets calculated when taking the amplitude ratios of 7 white noises in the range of 0.05 - 0.2 are respectively shown for the deep anesthesia state of a certain window and the awake state of a certain window, where there are 8 eiPDFs; the abscissa in the figure is the standard deviation distribution (d = -4:4), and the ordinate is the probability density; specifically, the green solid line is the eiPDF obtained by processing EEG data through the monitoring method of the present application, and the black dashed line represents the probability density function of the standard normal distribution, and they are both shown in the figure in the form of their respective probability density distribution curves.

[0133] To further quantify the modulation degree (i.e., the awake degree of the patient), subtract the probability density function of the standard normal distribution from the eiPDF of the EEG data (i.e., Figure 5 and Figure 7 subtract the black dashed line from the green solid line in Figure 6 and Figure 8 respectively), and the final set of intrinsic probability density functions (F-eiPDF) as shown in Figure 5 and Figure 7 is obtained. It is displayed in the form of a two-dimensional graph and can be used as the processing result of the current window. In each figure, the abscissa is each F-eiPDF corresponding to Figure 5 and Figure 7 (after subtraction), and the ordinate is the standard deviation distribution (d) corresponding to Figure 5 and Figure 7 . The differences (the differences between the eiPDF and the probability density function of the standard normal distribution) corresponding to 8 F-eiPDFs at different d are shown in the figure in different colors (representing different values).

[0134] Although the above steps describe how to obtain the initial set of intrinsic probability density functions (eiPDF) and its set of the current window, as well as the final set of intrinsic probability density functions (F-eiPDF) and its set; it is obvious that by using the above method, the eiPDF and its set and the F-eiPDF and its set of any window can be obtained respectively. Each eiPDF set includes multiple eiPDFs, for example, 8, and each F-eiPDF set includes multiple F-eiPDFs.

[0135] Output step of S600 data: Sort and output at least a part of at least one of the multiple initial intrinsic probability density functions (iPDF), multiple final intrinsic probability density functions (F-iPDF), multiple initial set of intrinsic probability density functions (eiPDF), and multiple final set of intrinsic probability density functions (F-eiPDF) of each window. The result is displayed on the monitor and archived, and can be used as a guide for the depth of clinical anesthesia.

[0136] In a specific embodiment, each iPDF in the iPDF set of each window is output to obtain a two-dimensional graph (i.e., iPDF monitoring graph) of each iPDF on the time axis. The number of graphs corresponds to the number of iPDFs, and these graphs can be used as monitoring guidance for clinical anesthesia.

[0137] In a preferred embodiment, each F-iPDF in the F-iPDF set of each window is output to obtain a two-dimensional graph (i.e., F-iPDF monitoring graph) of each F-iPDF on the time axis. The number of graphs corresponds to the number of F-iPDFs, and these graphs can be more preferably used as monitoring guidance for clinical anesthesia.

[0138] In a specific embodiment, each eiPDF in the eiPDF set of each window is output to obtain a two-dimensional graph (i.e., eiPDF monitoring graph) of each eiPDF on the time axis. The number of graphs corresponds to the number of eiPDFs, and these graphs can be used as monitoring guidance for clinical anesthesia.

[0139] In a preferred embodiment, each F-eiPDF in the F-eiPDF set of each window is output to obtain a two-dimensional graph (i.e., F-eiPDF monitoring graph) of each F-eiPDF on the time axis. The number of graphs corresponds to the number of F-eiPDFs, and these graphs can be more preferably used as monitoring guidance for clinical anesthesia.

[0140] In a specific embodiment, each iPDF in the iPDF set of each window is respectively output at the standard deviation distribution d to obtain a curve (i.e., iPDF monitoring curve) of a single iPDF at a single standard deviation distribution d on the time axis. The number of curves corresponds to the number of iPDFs, and these curves can be used as monitoring guidance for clinical anesthesia.

[0141] In a preferred embodiment, each F-iPDF in the F-iPDF set of each window is respectively output at the standard deviation distribution d to obtain a curve (i.e., F-iPDF monitoring curve) of a single F-iPDF at a single standard deviation distribution d on the time axis. The number of curves corresponds to the number of F-iPDFs, and these curves can be more preferably used as monitoring guidance for clinical anesthesia.

[0142] In a specific embodiment, each eiPDF in the eiPDF set of each window is respectively output at the standard deviation distribution d to obtain a curve (i.e., eiPDF monitoring curve) of a single eiPDF at a single standard deviation distribution d on the time axis. The number of curves corresponds to the number of eiPDFs, and these curves can be used as monitoring guidance for clinical anesthesia.

[0143] In a preferred embodiment, each F-eiPDF in the F-eiPDF set of each window is output separately at the standard deviation distribution d, obtaining a curve of a single F-eiPDF on a single standard deviation distribution d on the time axis (i.e., the F-eiPDF monitoring curve). The number of curves corresponds to the number of F-eiPDFs, and these curves can be more preferably used as the monitoring guidance for clinical anesthesia.

[0144] It should be understood that the two-dimensional monitoring map can be generated in ways such as pcolor map, contour map, etc.; they have the same values as the corresponding monitoring curves. In addition, when using the F-iPDF and F-eiPDF monitoring maps, since they are more intuitive, they can replace the iPDF and eiPDF monitoring maps; similarly, the F-iPDF and F-eiPDF monitoring curves can also replace the iPDF and eiPDF monitoring curves. In a preferred embodiment, the monitoring map and the monitoring curve are output simultaneously for monitoring guidance.

[0145] In specific operations, taking the output of the F-eiPDF set as the final result as an example, the output of the F-iPDF set (before set averaging) is similar. In step S100, the first raw data obtained from the first window is processed successively using steps S200, S300, S400, and S500, and the eiPDF set and F-eiPDF set of the first window can be obtained, which respectively include multiple eiPDFs and F-eiPDFs, assumed to be 8; the second raw data obtained from the second window is processed successively using steps S200, S300, S400, and S500, and the eiPDF set and F-eiPDF set of the second window can be obtained, which also respectively include multiple eiPDFs and F-eiPDFs, assumed to be 8; after the third raw data obtained from the third window is processed successively using steps S200, S300, S400, and S500, the eiPDF set and F-eiPDF set of the third window can be obtained, which also respectively include multiple eiPDFs and F-eiPDFs, assumed to be 8.... According to such a calculation method, during the data acquisition process, the newly acquired window data can be continuously calculated to obtain their respective eiPDF sets and F-eiPDF sets. In this example, the F-eiPDF set is taken as the object of description: each F-eiPDF (F-e1 PDF, F-e2PDF, F-e3PDF,..., F-e8PDF) in the F-eiPDF set of each window is arranged in the order of data acquisition time, so as to obtain the sequentially output F-eiPDF monitoring map and F-eiPDF monitoring curve. Specifically, the change trends of F-e1 PDF to F-e8PDF output in chronological order for all windows (the entire anesthesia operation process) can be displayed in an F-eiPDF monitoring map. Refer to Figure 9; It is also possible to separately display, among the 8 curves, the F-e1 PDF monitoring curves, F-e2 PDF monitoring curves, F-e3 PDF monitoring curves, F-e4 PDF monitoring curves, F-e5 PDF monitoring curves, F-e6 PDF monitoring curves, F-e7 PDF monitoring curves, and F-e8 PDF monitoring curves output in chronological order for all windows (the entire anesthesia operation process) with respect to a certain d, for reference Figure 10 and Figure 11 .

[0146] The following will specifically describe the F-ei PDF monitoring diagrams and F-ei PDF monitoring curves of an anesthesia operation process respectively, in order to more clearly describe and understand the present application Figures 9 - 11 in combination with the attached

[0147] Figure 9 shows the original data collected during the entire anesthesia operation process (within the time axis of 0 - 8700 s, taken as the abscissa), and the F-ei PDF monitoring diagrams respectively obtained through the monitoring method described in this embodiment, corresponding to F-e1 PDF to F-e8 PDF in the F-ei PDF sets of each window

[0148] Taking Figure 9 the upper F-e2 PDF in [figure] as an example, where the abscissa correspondingly represents the operation time from 0 to 8700 s, from the start to the end of the operation; the ordinate correspondingly represents the F-e2 PDF values of each time window (marked with different colors) within the range of d = -4:4; each F-e2 PDF value is obtained by using the original data of the current time window through steps S100 - S500. It can be seen from the figure that the F-e2 PDF values at the front stage of the monitoring diagram are relatively high (the colors tend to be blue and red with a large difference), and fluctuate severely, indicating that it is in the operation preparation and induction stage; then the F-e2 PDF values in the middle stage are relatively stable (the colors tend to be yellow and green with a small difference), indicating that it is in the middle stage of the operation; the final stage is the later stage of the operation, and the patient's consciousness gradually wakes up, so the F-e2 PDF values increase again and fluctuate severely

[0149] Figure 10 and Figure 11The original data collected within the time axis of 0 - 8700 s during the entire anesthesia surgery process is shown. Eight F - eiPDF monitoring curves for d = 0 and d = - 4 respectively obtained through the monitoring method described in this embodiment correspond to F - e1 PDF to F - e8PDF in the F - eiPDF set of each window. Among them, the abscissa of each monitoring curve is the time axis, and the ordinate is F - eiPDF (the difference between the eiPDF at this d and the probability density function of the standard normal distribution). That is, each F - eiPDF corresponding to each second in 8700 s is the result obtained by calculating 6000 original data through the steps described above. Taking Figure 11 the second figure in Figure 12c (also Figure 10 and Figure 11 ) as an example, where the abscissa correspondingly is the surgery time of 0 - 8700 s, from the start to the end of the surgery; the ordinate correspondingly is the F - e2PDF value of each time window; each F - e2PDF value is obtained by using the original data of the current time window through steps S100 - S500. It can be seen from the figure that the F - e2PDF value at the front section of the monitoring curve is relatively high and fluctuates severely, indicating the surgery preparation and induction stage; then the F - e2PDF value in the middle section is relatively stable, indicating the middle stage of the surgery; the final section is the late stage of the surgery, and the patient's consciousness gradually wakes up, so the F - e2PDF value rises again and fluctuates severely. It can be understood that Figure 10 and Figure 11 The monitoring curves shown are Figure 9 part of the results in Figure 10 For example, Figure 9 are all the values when d = 0 in Figure 11 and Figure 9 are all the values when d = - 4 in Figure 5 and Figure 7 They are mainly different in the manifestation form and are all obtained by processing similar to Figure 5 and Figure 7 . Relatively speaking, the performance of the monitoring graph is more perfect (including d = - 4:4), the monitoring curve is more intuitive, and the two can be used alone or together to assist each other.

[0150] Each ensemble empirical probability density function in the ensemble empirical probability density function set can be used as a monitoring parameter. When output on the monitor, all can be output. For example, all F - e1 PDF to F - e8PDF shown in Figure 9 are output, or Figure 10 and Figure 11 The 8 monitoring curves (regarding d = 0 and d = - 4) shown in or a part of them. In some embodiments, since the ensemble empirical probability density function depends on the scale of the empirical mode function, and through experience and data comparison, it is found that the F - e3PDF (d = 0) monitoring curve in the final ensemble empirical probability density function (F - eiPDF)Figure 10 the third figure in Figure 11 and the F-e2PDF(d = -4) monitoring curve (the second figure in

[0151] Figure 12b and Figure 12c are EEG signals from the same source. The F-e3PDF(d = 0) monitoring curve and the F-e2PDF(d = -4) monitoring curve obtained by the monitoring method of this application are the enlarged views of the F-e3PDF(d = 0) monitoring curve and the F-e2PDF(d = -4) monitoring curve in Figure 10 and Figure 11 . The characteristics they show are close to the trend of the corresponding BIS curve ( Figure 12a ), indicating that the monitoring method of this application is reliable.

[0152] Comparative analysis:

[0153] This content compares the anesthesia depth monitoring method based on the ensemble empirical probability density function provided by this application with the BIS anesthesia monitoring method in the prior art.

[0154] Both monitoring methods use the same original data. Among them, Figure 12a is the monitoring curve obtained by using the BIS monitoring method (hereinafter referred to as the BIS monitoring curve) and the corresponding preprocessed EEG data. Figure 12a In Figure 12b and 12c are the F-e3PDF(d = 0) monitoring curve and the F-e2PDF(d = -4) monitoring curve obtained in this embodiment.

[0155] Figure 12a shows the recorded EEG signal and its BIS value, where the sharp changes during induction are obvious. The data was disturbed during the operation, which led to a great interference in the BIS score. The BIS score also shows that in the later stage of the operation, the recovery time of the patient was prolonged.

[0156] However, the final ensemble empirical probability density function (F-eiPDF) results are different. During wakefulness, the F-eiPDF is significantly super-Gaussian distributed, as shown in Figure 7 , 8As shown, this indicates that the brain is fully functional and all interactions lead to modulation, thus presenting a super-Gaussian distribution. However, after anesthesia induction, the interconnections and interactions are blocked; therefore, during surgery, the F-eiPDF is almost Gaussian distributed, as shown in Figure 5 and 6 . This embodiment uses a sliding window to give nearly real-time scores of the F-eiPDF and construct a surgical history, as shown in Figure 12b and 12c ; the results are similar to the BIS in some aspects but different in some aspects. The similarities lie in the overall characteristics. The F-eiPDF also gives the profile of the anesthesia curve from induction to recovery; the induction period is obvious. The key difference is that there are no large perturbations during the mid-surgery period because the patient does not wake up; minor perturbations are possible. The most significant difference is during the recovery process; in most surgeries, the recovery is as sudden as the induction; the BIS fails to show this feature, but the F-eiPDF monitoring curve has large fluctuations in the later stage of surgery, clearly showing the changes in the patient's consciousness. Therefore, it can provide better monitoring guidance for doctors.

[0157] This embodiment provides a new tool for quantitatively studying anesthesia; since this method measures consciousness, it is independent of the anesthetic used, the patient's age, and gender; it can be used as a general method for measuring anesthetic depth. In summary, we believe that with the theoretical breakthrough of the ensemble eigenprobability density function, we have a new tool for quantitatively studying anesthesia. Although the eiPDF and F-eiPDF are phenomenological, for clinical applications, the proposed method will take a big step forward from the current state-of-the-art technology and meet many urgent needs.

[0158] As mentioned above, the monitoring method provided by this application can also be used for monitoring muscle activity. Figure 13 is the preprocessed electromyogram data of the tremor muscle of a Parkinson's patient, Figure 14 is the F-e3PDF(d = 0) monitoring curve obtained by processing the electromyogram data of each window using the monitoring method provided by this application, Figure 15 is the F-e2PDF(d = -4) monitoring curve obtained by processing the electromyogram data of each window using the monitoring method provided by this application, Figure 16 is the F-eiPDF monitoring map obtained by processing the electromyogram data of each window using the monitoring method provided by this application. Figure 17 and Figure 18 are respectively 8 F-eiPDF monitoring curves for d = 0 and 8 F-eiPDF monitoring curves for d = -4, Figure 14 and Figure 15Select from them. The sampling frequency of the electromyography data is 1000 Hz, the window length is set to 10 s, and the sliding step is 1 s. After completing the preprocessing steps for the electromyography data within the window, it enters the same or similar steps as the anesthesia depth monitoring to obtain Figures 16 - 18 the monitoring graph and monitoring curve shown. The muscle tremor activity of Parkinson's patients is the action potential with a fixed rhythm sent from the upper motor neurons. The F-eiPDF monitoring graph and monitoring curve can reflect the modulation intensity of the upper neurons on the muscles during the tremor of the Parkinson's patient. Through monitoring for a period of time, the tremor occurrence structure (percentage of no tremor, percentage of tremor, etc.) of the Parkinson's patient within this period can be quantified. It can be seen that the monitoring method based on the intrinsic probability density function is also particularly applicable in muscle activity monitoring.

[0159] The second embodiment of this application provides an anesthesia depth monitoring system based on the intrinsic probability density function (which can be abbreviated as the monitoring system), and this system is used to implement the monitoring method described in any of the previous embodiments. The monitoring system includes: a data acquisition module and a data analysis module, where,

[0160] The data acquisition module is configured to collect electroencephalogram data using electrodes to obtain the original data;

[0161] The data analysis module includes a windowing module, a data preprocessing module, an EEMD decomposition module, an iPDF calculation module, an eiPDF calculation module, and a result output module; where,

[0162] The windowing module is configured to window the collected electroencephalogram data to obtain the original data of each window;

[0163] The data preprocessing module is configured to filter, remove outliers, and remove electrical noise from the collected original data to obtain the data time series to be analyzed;

[0164] The EEMD decomposition module is configured to adaptively decompose the data time series to be analyzed in the current window into multiple E-IMFs using the EEMD analysis method;

[0165] The iPDF calculation module is configured to select the first N E-IMFs from multiple E-IMFs, use formula (1) to find the partial component sums to obtain N partial component sums, and then calculate the probability density function of each partial component sum, so as to obtain the N iPDFs of the current window, forming an iPDF set; formula (1) refers to the first embodiment;

[0166] The eiPDF calculation module is configured to set white noise sequences with multiple different amplitude ratios, apply them to the EEMD decomposition module and the iPDF calculation module respectively, and obtain multiple rounds of iPDF sets; take the set average value of the iPDF sets in multiple rounds as the result to obtain multiple initial ensemble intrinsic probability density functions (eiPDFs) of the current window, and form an eiPDF set; and

[0167] The result output module is configured to sort and output multiple initial ensemble intrinsic probability density functions (eiPDFs) of each window.

[0168] In some embodiments, the iPDF calculation module is further configured to subtract the probability density function of the standard normal distribution from the obtained multiple iPDFs respectively to obtain multiple final intrinsic probability density functions (F-iPDFs), and form an F-iPDF set;

[0169] The eiPDF calculation module is further configured to subtract the probability density function of the standard normal distribution from the obtained multiple eiPDFs respectively to obtain multiple final ensemble intrinsic probability density functions (F-eiPDFs), and form an F-eiPDF set;

[0170] The result output module is further configured to sort and output multiple final intrinsic probability density functions (F-iPDFs) and / or final ensemble intrinsic probability density functions (F-eiPDFs) of each window.

[0171] In some embodiments, the monitoring system may further include:

[0172] The data archiving module is configured to store data for future reference;

[0173] The data display module is configured to display the output result of the result output module on a monitor, which can be used as a guide for the depth of clinical anesthesia.

[0174] It should be understood that the technical features described in the monitoring method can also be reasonably applied to the monitoring system of this embodiment, and those that have been described will not be repeated. It should be noted that the above-mentioned each module can be a functional module or a program module, and can be implemented either by software or by hardware. For the modules implemented by hardware, the above-mentioned each module can be located in the same processor, or the above-mentioned each module can also be located in different processors in any combination form.

[0175] The third embodiment of the present application provides an application of the monitoring method described in any of the previous embodiments, which can be used for medical monitoring. The medical monitoring includes but is not limited to anesthesia depth monitoring and muscle activity monitoring.

[0176] The fourth embodiment of the present application provides a computer device, which may include a processor 101 and a memory 102 storing computer program instructions, as Figure 19 shown.

[0177] Specifically, the processor 101 may include a central processing unit (CPU), or an application specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present application. The memory 102 may include a mass storage for data or instructions. By way of example and not limitation, the memory 102 may include a hard disk drive (HDD), a floppy disk drive, a solid state drive (SSD), a flash memory, an optical disc, a magneto-optical disc, a magnetic tape, or a universal serial bus (USB) drive, or a combination of two or more of these. In a suitable case, the memory 102 may include a removable or non-removable (or fixed) medium. In a suitable case, the memory 102 may be internal or external to the data processing device.

[0178] The memory 102 may be used to store or cache various data files required for processing and / or communication use, as well as possible computer program instructions executed by the processor 101.

[0179] The processor 101 reads and executes the computer program instructions stored in the memory 102 to implement the monitoring method described in any of the foregoing embodiments.

[0180] In some embodiments, the computer device may further include a communication interface 103 and a bus 104. Among them, as Figure 19 shown, the processor 101, the memory 102, and the communication interface 103 are connected through the bus 104 and complete communication with each other. The communication interface 103 is used to implement communication between the modules, devices, units, and / or devices in the embodiments of the present application. The communication interface 103 may also implement data communication with other components.

[0181] The fifth embodiment of the present application provides a computer-readable storage medium, on which computer program instructions are stored; when the computer program instructions are executed by a processor, the monitoring method described in any of the foregoing embodiments is implemented.

[0182] The technical features of the above-described embodiments can be combined arbitrarily. For the sake of brevity of description, not all possible combinations of the technical features in the above-described embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as within the scope described in this specification.

[0183] The above-described embodiments merely represent several embodiments of the present application, and their descriptions are relatively specific and detailed, but they should not be construed as limiting the scope of the application. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present application, several modifications and improvements can still be made, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the appended claims.

Claims

1. A monitoring method based on intrinsic probability density function, characterized in that: The following steps are involved: Steps to obtain raw data: collect data as raw data; Data preprocessing step: preprocess the acquired raw data to obtain the data time series to be analyzed; The steps of applying EEMD to the data time series are as follows: using the EEMD analysis method to process the data time series and adaptively decomposing it into a plurality of ensemble intrinsic mode functions (E-IMFs); and The calculation steps of the intrinsic probability density function (the calculation steps of iPDF): Formula (1) is used to calculate the partial component sum of N set intrinsic mode functions to obtain N partial component sums; Among them, c j (t) is the collective intrinsic mode function (E-IMF), t represents the number of data, N represents the total number of E-IMFs, which is a positive integer, n represents the nth E-IMF, and x n (t) represents the sum from the first E-IMF to the nth E-IMF; Then, the probability density function of the sum of each partial component is calculated to obtain N initial intrinsic probability density functions (iPDFs), which constitute a set of initial intrinsic probability density functions (iPDF set); At least a portion of the multiple iPDFs are used as output of the monitoring method.

2. The monitoring method according to claim 1, characterized in that: In the iPDF calculation step, further, the probability density functions of the standard normal distribution are subtracted from the obtained multiple iPDFs to obtain multiple final intrinsic probability density functions (F-iPDFs), and at least a part of the multiple F-iPDFs is used as the output of the monitoring method, which can replace the output of the iPDF.

3. The monitoring method according to claim 1, characterized in that: The method further comprises a step of calculating a set intrinsic probability density function (eiPDF calculation step): setting a plurality of white noise sequences with different amplitude ratios, applying them to the step of applying the EEMD to the data time series, respectively, and performing the iPDF calculation step to obtain multiple rounds of iPDF sets; taking the set average of the iPDF sets under multiple rounds as the result, obtaining a plurality of initial set intrinsic probability density functions (eiPDF); At least a part of the plurality of eiPDFs can serve as output of the monitoring method and can replace the output of the iPDF.

4. The monitoring method according to claim 3, characterized in that: In the calculation step of the eiPDF, further, the probability density function of the standard normal distribution is subtracted from the obtained multiple eiPDFs to obtain multiple final collective intrinsic probability density functions (F-eiPDFs), and at least a part of the multiple F-eiPDFs is used as the output of the monitoring method, which can replace the output of eiPDF.

5. The monitoring method according to any one of claims 1 to 4, characterized in that: In the step of obtaining the raw data, a sliding window is used to collect data, and the data of each window is used as the raw data; the output of each window can be obtained by using the monitoring method; The monitoring method also includes a data output step, wherein the output of each window is displayed as a result on a monitor and archived as a monitoring guide.

6. The monitoring method according to claim 5, characterized in that: In the step of outputting the data, at least one of the following methods is selected: (1) Output all or part of the multiple iPDFs in each window to obtain an iPDF monitoring diagram of each iPDF on the time axis for monitoring guidance; (2) Outputting all or part of the multiple F-iPDFs of each window to obtain an F-iPDF monitoring diagram of each F-iPDF on the time axis for monitoring guidance; (3) Output all or part of the multiple eiPDFs of each window to obtain an eiPDF monitoring diagram of each eiPDF on the time axis for monitoring and guidance; (4) Outputting all or part of the multiple F-eiPDFs of each window to obtain an F-eiPDF monitoring diagram of each F-eiPDF on the time axis for monitoring guidance; (5) Outputting all or part of the multiple iPDFs of each window in the standard deviation distribution d, respectively, and obtaining an iPDF monitoring curve of a single iPDF in the single standard deviation distribution d on the time axis for monitoring guidance; (6) Outputting all or part of the multiple F-iPDFs of each window in the standard deviation distribution d, respectively, to obtain an F-iPDF monitoring curve on the time axis about a single F-iPDF in the single standard deviation distribution d, for monitoring guidance; (7) Outputting all or part of the multiple eiPDFs of each window in the standard deviation distribution d respectively, and obtaining an eiPDF monitoring curve of a single eiPDF in a single standard deviation distribution d on the time axis for monitoring and guidance; (8) All or part of the multiple F-eiPDFs of each window are outputted in the standard deviation distribution d, respectively, to obtain the F-eiPDF monitoring curve of a single F-eiPDF in the single standard deviation distribution d on the time axis for monitoring guidance.

7. The monitoring method according to claim 5, characterized in that: In the data output step, the third monitoring curve in F-eiPDF when d=0, i.e., F-e3PDF (d=0), and the second monitoring curve in F-eiPDF when d=-4, i.e., F-e2PDF (d=-4), are selected as the main results, and the other monitoring curves are used as supporting results.

8. The monitoring method according to any one of claims 1 to 4, 6 and 7, characterized in that: In the step of acquiring raw data, the sampling frequency of the data is not less than 100 Hz; in the step of preprocessing the data, the preprocessing includes filtering, outlier removal and electrical noise removal.

9. The monitoring method according to any one of claims 1 to 4, 6 and 7, characterized in that: The step of applying EEMD to the data time series specifically includes: Add white noise: add a white noise sequence to the data time series to be analyzed; the amplitude ratio of the white noise is any value between 0.05 and 0.2; Decompose IMFs using EMD: Adaptively decompose the data time series with white noise into multiple initial intrinsic mode functions (IMFs) using EMD; Repeat the steps of adding white noise and decomposing IMFs using EMD, but use a newly randomly generated white noise sequence with the same amplitude ratio each time; thereby obtaining multiple batches with multiple IMFs respectively; The ensemble average of multiple IMFs of multiple batches is taken as the decomposition result to obtain the multiple E-IMFs.

10. The monitoring method according to claim 9, characterized in that: In the step of applying the EEMD to the data time series, the step of adding white noise and the step of decomposing IMFs using EMD are repeated 10-20 times to obtain 10-20 batches; in the step of calculating the iPDF, a maximum of 8 of the multiple E-IMFs are selected for calculating the sum of partial components; that is, N is a positive integer less than or equal to 8; in the step of calculating the eiPDF, the amplitude ratio of the white noise is set to 0.05-0.2, and 6-16 sets are taken for trial as the white noise sequences with multiple different amplitude ratios.

11. A monitoring system based on an intrinsic probability density function, characterized in that: It includes data acquisition module and data analysis module, among which, The data acquisition module is configured to collect data, thereby obtaining raw data; The data analysis module includes a windowing module, a data preprocessing module, an EEMD decomposition module, an iPDF calculation module and a result output module; wherein, The windowing module is configured to collect data in a windowed manner to obtain raw data of each window; The data preprocessing module is configured to filter, remove outliers and remove electrical noise from the collected raw data to obtain a data time series to be analyzed; The EEMD decomposition module is configured to adaptively decompose the data time series to be analyzed in the current window into multiple E-IMFs using the EEMD analysis method; The iPDF calculation module is configured to select the first N E-IMFs from the plurality of E-IMFs, calculate the partial component sums using formula (1), and obtain N partial component sums. Among them, c j (t) is the E-IMF, t represents the number of data, N represents the total number of E-IMFs, which is a positive integer, n represents the nth E-IMF, x n (t) represents the sum from the first E-IMF to the nth E-IMF; Then, the probability density function of the sum of each partial component is calculated to obtain the iPDF set of the current window; The result output module is configured to sort and output the multiple iPDFs in the iPDF set of each window.

12. The monitoring system according to claim 11, characterized in that: Further comprising: an eiPDF calculation module, configured to set a plurality of white noise sequences with different amplitude ratios, apply them to the EEMD decomposition module and the iPDF calculation module respectively, and obtain multiple rounds of iPDF sets; taking the set average of the iPDF sets under multiple rounds as the result, and obtaining multiple eiPDFs of the current window; and The result output module is configured to output the multiple eiPDFs of each window in order to replace the output of the iPDF set.

13. The monitoring system according to claim 12, characterized in that: The eiPDF calculation module is further configured to respectively subtract the probability density function of the standard normal distribution from the obtained multiple eiPDFs to obtain multiple F-eiPDFs; The result output module is further configured to output the multiple F-eiPDFs of each window in order to replace the output of eiPDF.

14. The monitoring system according to any one of claims 11 to 13, characterized in that: Also includes: a data archiving module configured to store data for future reference; The data display module is configured to display the output result of the result output module on a monitor as a monitoring guide.

15. An application of the monitoring method based on the intrinsic probability density function according to any one of claims 1 to 10, wherein the monitoring method can be used for medical monitoring, including anesthesia depth monitoring and muscle activity monitoring.

16. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the monitoring method based on the intrinsic probability density function as described in any one of claims 1 to 10 when executing the computer program.

17. A computer-readable storage medium having a computer program stored thereon, wherein when the program is executed by a processor, the monitoring method based on the intrinsic probability density function as described in any one of claims 1 to 10 is implemented.

Images